Environmental Engineering Reference
In-Depth Information
dc
(
t
)
1
V (
=
f
(
t
)−
q s
(
t
)
c s
(
t
)−(
p
(
t
)−
q sp
(
t
)
c
(
t
))),
(5.32)
dt
where q s
(
t
)
is the volumetric feed flowrate defined as q s
(
t
)=
q
(
t
)+(
f
(
t
)/
ρ s
)
with
ρ s being density of the ore. Volumetric flowrate q sp
is fixed by the pump. The
level of the slurry can then be calculated as a function of the volume.
Ore hardness is assumed to be transported trough the sump as
(
t
)
(
)
(
)
d γ
t
q
t
=
) (
γ f
(
t
)−
γ
(
t
)).
(5.33)
dt
V
(
t
5.5 Model-based Control Systems Design
This section illustrates the use of a simulator in the design of model-based control
strategies for different processes. According to classical design procedure, the major
steps for an advanced model-based controller using a simulator can be summarized
as follows:
1. Define objectives. Identify control objectives with respect to economics and pro-
cess behavior (including safety).
2. Variables identification. Identify manipulated variables, disturbances and con-
trolled variables based on control objectives, process knowledge and experience.
Identify possible restrictions.
3. Experimental testing. Perform plant tests to obtain suitable data for dynamic
modeling. If possible, measure disturbance variables as well. Collect data for
further analysis.
4. Model identification. Identify a linear empirical model based on the gathered
data. If the fit is not good enough, the model can be improved by seeking non-
linear transformations on selected variables. This can be guided by using phe-
nomenological knowledge. After fitting the model, it is always useful to perform
some kind of validation; i.e. , to analyze the static and dynamic characteristic of
the model and check whether they are consistent with conditions at the plant.
5. Design. Design the final controller based on the identified model. In some cases,
an identified model can be used to tune a pre-defined controller such as a PID, or
can be an integral part of the controller.
6. Closed loop simulation. Perform off-line testing and tuning of the proposed con-
troller by testing it in a closed loop against the nonlinear simulated process
model.
After finishing the design stage, an implementation and commissioning stage
follows, with final tuning adjustments made on-line. In the following examples, the
simulator will be used to carry out most of the steps described above.
Search WWH ::




Custom Search